an enthusiastic austrian solaris user

Wine on OpenIndiana - LiberOffice ok, Skype fails :-(

2012-01-28T10:41:00.000+01:00

Because of the current step on trying to get LibreOffice to OpenIndiana,
I read in a mailing list about using LibreOffice 3.5rc2 on wine.

The wine Package itself is in the repository sfe-encumbred
( http://pkg.openindiana.org/sfe-encumbered/ ).

Wine needs also the gecko engine for html rendering.

The simple commands were:

$ wget http://downloads.sourceforge.net/wine/wine_gecko-1.4-x86.msi
$ sudo mkdir -p /usr/share/wine/gecko
$ sudo mv wine_gecko-1.4-x86.msi /usr/share/wine/gecko/

from http://wiki.winehq.org/Gecko

LibreOffice is from the link embedded in the mailinglist entry.

It is a msi package, which has to be installed with

wine msiexec /i xyz.msi

I also tried Skype, but the latest version has hugh problems and does
not start.

LibreOffice works so far.

But with proxies I have problems.

UrukDroid 1.6.1 -> Flight-Mode -> resume fails

2012-01-26T10:44:00.002+01:00

My Archos-Tablet 10.1 IT is on UrukDroid 1.6.1.

I'm using a Widget for activating WiFi.

The last time I attempted to use the Flight-Mode switch, but as it seems, every
2nd wakeup does not wakeup the screen, so I had to do a hard reset ( 10 secs on the power button ) and reboot.

Without using the Flight-Mode and only activate/deactivate the WIFI resume works
for me every time.

Hidden or for me unknow NetBeans Feature to show two files side by side

2012-01-26T10:41:00.001+01:00

In NetBeans I'd sometimes compare the some code in the same file side by side.

To get the same file opened a second time, there is a clone-Item in the right-click-menu on the tab-Item.

To see it side by side, I dragged one around to fit it side by side, but the
right-click-menu has a "New Document Tab Group". This does it automatically and
has also a "Collapse Document Tab Group".

Problems on OpenIndiana 151a with the g++ from the SFE repository

2012-01-13T18:09:00.000+01:00

I'm trying to compile a simple C++ file on my OpenIndiana 151a with
the g++ from the SFE repository.

root@solewer:/usr/gcc/4.6# g++ --version
g++ (GCC) 4.6.2
Copyright (C) 2011 Free Software Foundation, Inc.
This is free software; see the source for copying conditions. There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.

But this results in a lot of errors during the assembly stage:

root@solewer:~# g++ a.cpp
Assembler: a.cpp
    "/var/tmp//cclAa4ot.s", line 80 : Syntax error
    Near line: "    .uleb128 .LLSDACSE549-.LLSDACSB549"
    "/var/tmp//cclAa4ot.s", line 82 : Syntax error
    Near line: "    .uleb128 .LEHB0-.LFB549"
    "/var/tmp//cclAa4ot.s", line 83 : Syntax error
    Near line: "    .uleb128 .LEHE0-.LEHB0"

With the g++ from /usr/sfw/bin ( pkg://openindiana.org/developer/gcc-3
) the compilation succeeded.

I've now started with execsnoop to see whats happening:

g++ from the SFE repository uses the default assembler:

UID    PID   PPID ARGS
    0   7755   7662 g++ a.cpp
    0   7756   7755
/usr/gcc/4.6/lib/gcc/i386-pc-

solaris2.11/4.6.2/cc1plus -quiet a.cpp
-quiet -dum
0 7758 7662 /usr/bin/hostname
0 7757 7755 /usr/ccs/bin/as -s -o /var/tmp//ccYHayjp.o
/var/tmp//ccXHayjp.s

but using /usr/sfw/bin/g++ from (
pkg://openindiana.org/developer/gcc-3 ) the gcc has its own assembler

    0   7762   7662 /usr/sfw/bin/g++ a.cpp
    0   7763   7762
/usr/sfw/libexec/gcc/i386-pc-solaris2.11/3.4.3/cc1plus -quiet a.cpp
-quiet -dum
    0   7764   7762 /usr/sfw/libexec/gcc/i386-pc-solaris2.11/3.4.3/as
--traditional-format -Qy -s -
    0   7765   7762
/usr/sfw/libexec/gcc/i386-pc-solaris2.11/3.4.3/collect2 -Y
P,/usr/ccs/lib:/lib:

Compiling a C file with gcc works for both 4.6.2 and 3.4.3

The usage of the assembler under ccs is the problem. I searched for
other assemblers on my system and found:

root@solewer:/usr# find . -name as -print
./gnu/i386-pc-solaris2.11/bin/

as
./bin/as
./sfw/libexec/gcc/i386-pc-solaris2.11/3.4.3/as
./ccs/bin/as
./share/locale/as

But the searchpath of gcc does not search in /usr/gnu:

[ewer@solewer:~/tmp]$g++ -print-search-dirs
...
programs: =/usr/gcc/4.6/lib/gcc/i386-pc-solaris2.11/4.6.2/:/usr/gcc/4.6/lib/gcc/i386-pc-solaris2.11/4.6.2/:/usr/gcc/4.6/lib/gcc/i386-pc-solaris2.11/:/usr/gcc/4.6/lib/gcc/i386-pc-solaris2.11/4.6.2/:/usr/gcc/4.6/lib/gcc/i386-pc-solaris2.11/:/usr/gcc/4.6/lib/gcc/i386-pc-solaris2.11/4.6.2/../../../../i386-pc-solaris2.11/bin/i386-pc-solaris2.11/4.6.2/:/usr/gcc/4.6/lib/gcc/i386-pc-solaris2.11/4.6.2/../../../../i386-pc-solaris2.11/bin/:/usr/ccs/bin/i386-pc-solaris2.11/4.6.2/:/usr/ccs/bin/
...

Programs are only search in the /usr/gcc/4.6 folders. So I made a soft
link from

/usr/gcc/4.6/i386-pc-solaris2.11 -> /usr/gnu/i386-pc-solaris2.11/

After this link the compilation and linking works again without
problems and the
created executeables can run.

But I don't know why or if this has changed. Because I had already compiled

with this compiler from the SFE repository.

Show sender of e-mail in thunderbird message pane

2012-01-10T12:56:00.002+01:00

Thunderbird shows in the message pane the "From".

I get also e-mails, where the senders are using the same name in From, but have
different entries in the sender tag.

To see also the sender in the Messages pane, you have to activate it in the config
editor.

Edit -> Preferences -> Advanced

Here the tab General contains a button "Config Editor"

Search for

mailnews.headers.showSender

and set it to true.
After a restart, thunderbird shows also the sender in the message pane.

Browser Testframework SST using Selenium

2012-01-04T13:15:00.000+01:00

For longer time I know about Selenium, which supports multiple language
to create testcases.
Some days ago, I found a link about SST -
SST - Web Test Framework

in the following blogpost from Corey Goldberg
This allows to simplify the testing, which are processed as
unit tests.

A simple example:

mytest/a.py:

from sst.actions import *

go_to( "http://www.ubuntu.com" )
assert_title_contains( "Homepage | Ubuntu" )

Execute:

[ewer@solewer:~/tmp]$sst-run -b Firefox -d mytest/
----------------------------------------------------------------------
starting SST...

date time: 2012-01-04 11:46
test directory: 'mytest/'
report format: 'console'
browser type: 'Firefox'
javascript disabled: False
shared directory: None
screenshots on error: False
failfast: False
debug: False
headless xserver: False

1 test cases loaded

----------------------------------------------------------------------
test_a (sst.runtests.TestA) ...
    Starting Firefox:
    Going to... http://www.ubuntu.com
    Waiting for get_element
    Stopping browser
ok

----------------------------------------------------------------------
Ran 1 test in 4.885s

OK
----------------------------------------------------------------------
[ewer@solewer:~/tmp]$

This opens a firefox windows, executes the task and shows the test.

lightning Plugin 1.1 for Thunderbird 9.x on OpenIndiana/Solaris is available

2012-01-04T10:06:00.000+01:00

Thunderbird for Solaris/ OpenIndiana x86 is available since some time, but I'm
also using lightning ( the calendar plugin ), which is not running, even when
patching the version in install.rdf.

But today I found a new plugin in the download folder for 1.1 which is
compatible with Thunderbird 9.0.
But currently there is also thunderbird 9.0.1 available, which I have running
here on OpenIndiana and the lightning plugin doesn't work.
But just unzip the xpi-Archiv, patch the maxVersion in install.rdf from 9.0
to 10.0 and you can install the plugin also on version 9.0.1 of thunderbird
and it shows the calendar.

Update to UrukDroid 1.6 on Archos Tablet IT 10.1 from RC7

2012-01-04T09:42:00.001+01:00

I upgraded my table from UrukDroid 1.6 RC7 to UrukDroid 1.6 ( RapidShare ).

The Upgrade itself ( Copy of Updatefile to /data/UrukUpdate ).

The update had no problems and everyting seems to run fine, but the forum
showed that the root account is missing.

After a reboot I had the same problem.

I solved it after a hint from sauron in the forum.

I switched to the recovery menu ( during reboot press the volume button when
the archos logo disappears )

Here I restored the rootfs-Backup, booted into UrukDroid 1.6 RC7 and started
SSH during boot, rebooted once to confirm a running sshd and installed
the updated again.
After the update the su to root is not working again, as expected, but now the
SSHD is running.

After logging in via ssh from remote with the default password "UrukDroid",
I could change the permission of /system/bin/su with

chmod ug+s /system/bin/su

and the

su -

in the Terminal works again.

New UrukDroid 1.6 RC7 is available

2011-12-30T19:32:00.000+01:00

Have updated my tablet to UrukDroid 1.6 RC7 - if no problems are found, this will become the
final release.

I upgraded and didn't experience any problems so far.

Compiling Qt 4.8.0 on OpenIndiana 151

2011-12-28T10:32:00.003+01:00

Procedure:

download Qt 4.8.0
Set Path to include the compiler ( /opt/SUNWspro/bin )
adjust the compiler settings to include the X11 libraries and include
qt-everywhere-opensource-src-4.8.0/mkspecs/solaris-cc/qmake.conf
QMAKE_INCDIR_X11 = /usr/X11/include
QMAKE_LIBDIR_X11 = /usr/X11/lib
QMAKE_INCDIR_OPENGL = /usr/X11/include
QMAKE_LIBDIR_OPENGL = /usr/X11/lib
configure:
CC=/opt/SUNWspro/bin/cc CXX=/opt/SUNWspro/bin/CC CXXFLAGS="-library=stlport4" LDFLAGS="-library=stlport4" ./configure --prefix=/tools/qt/4.8.0 -opensource
gmake
qfiledialog.cpp:886 needed the code from the else part of the preprocessor
helloconcurrent.cpp failed to compile ( it is from the examples )
had to do some tricking with the including of qfuture.h and the
path to src/corelib/concurrent/qfuture.h from include/QtCore/qfuture.h
I remove QFuture from helloconcurrent.cpp and the whole compile succeede, with some adjustments before.

Upgrade 1.6RC6 of UrukDroid for Archos Tablet available

2011-12-27T15:25:00.002+01:00

The next RC of UrukDroid 1.6 is available. -> UrukDroid 1.6RC6 .

Download and installation were no problems. With RC5 the battery usage seems to be faster.
The short minutes I've upgraded ( copy archive to /data/UrukUpdate ) didn't
show any problems.

2011-12-22T15:53:00.001+01:00

Yesterday, I upgraded my Archos InternetTablet 10.1 from UrukDroid 1.6RC4
to
UrukDroid 1.6RC5

Did a backup before and installed it. After the installation the automatic
reboot ran in a loop.
I had to do a hard reset, but afterwards the system booted without problems.

reduce system usage of firefox by Add-On ToManyTabs

2011-12-21T10:24:00.000+01:00

Today I have read in the Book DTrace about an investigation about the
CPU usage of the firefox browser.
There was the hint about the Plugin ToManyTabs, which allows to move
open tabs to different rows, which act like closed tabs and the
CPU usage drops.

There is a small row in the header where you can switch between different
groups.
When you select an item it is moved to the open tabs and removed from the
group, but you can drag&drop it back to the row.

VirtualBox 4.1.8 update on OpenIndiana 151

2011-12-20T17:26:00.001+01:00

VirtualBox has released a new version 4.1.8 for Solaris.

Install on OpenIndiana was without problems, but still need LD_NODIRECT=1
for starting, otherwise VirtualBox fails or the Manager doesn't find the
virtual machines.

code-analyzer in Oracle Solaris Studio 12.3

2011-12-16T22:20:00.001+01:00

Oracle has release Solaris Studio 12.3 .
The suite has a new tool called code-analyzer.

It can do static code analyze by adding a compile directive -xanalyze=code.

I've used it for a project, but failed at first, because I had also to give the option on the linking step, but afterwards the executeable get a directory
called <executeablename>.analyze and here is a screenshot.

I had to mask a lot of code, because it's proprietary.
But the tool finds errors, although the code is running since years on different operating systems.

Something strange happend to my Crossbow Network

2011-12-16T11:05:00.000+01:00

I'm using the virtual network Crossbow on my OpenIndiana, to route the
traffice between Windows XP in VirtualBox, a Zone and the outer world by
using a virtual network switch.

It worked till Wednesday, but after shut down and starting on Thursday
XP could not connect anymore????.

The Zone can connect and can be pinged from the host, XP not????.

Now I've change to use NAT in VirtualBox and changed in Windows XP the
network to DHCP and the connection itself works but not over the virtual
network.

Have to investigate sometime later why this happened.

Update:
Found the problem in my own former post.

The problem seems that the mac adresse didn't match between VirtualBox
and the setting of the vnic as shown be dladm show-vnic

root@solewer:~# dladm show-vnic
LINK         OVER         SPEED MACADDRESS        MACADDRTYPE         VID
vnic1        vswitch1     0      2:8:20:55:9e:bf   random              0
vnicvsop0    vswitch1     0      2:8:20:7b:84:43   random              0
vnicsuse1    vswitch1     0      2:8:20:b3:bf:c1   random              0
vnicwinxp2   vswitch1     0      2:8:20:17:e7:c8   random              0

log4cplus && SysLogAppender

2011-12-12T16:41:00.003+01:00

I'm a little experimenting with logging frameworks for C++.
log4cplus is a nice one.

Now I'm doing the first test on OpenIndiana.

The ConfigureAndWatchThread does it's job.
When the app runs and afterwards the config file is copied into the
target directory, the logging starts.

But now I've configured the SysLogAppender:

log4cplus.appender.syslog=log4cplus::SysLogAppender
log4cplus.appender.syslog.ident=WARN
log4cplus.appender.syslog.facility=local0
log4cplus.appender.syslog.layout=log4cplus::PatternLayout
log4cplus.appender.syslog.layout.ConversionPattern=[%d{%d %m %Y %H:%M:%S}] [%t] %-5p %c{2} %%%x%% - %m [%l]%n

And when running the app with the corresponding setting publishes
the message into the configured syslog file:

Dec 12 16:37:44 solewer WARN: [ID 702911 local0.crit] [12 12 2011 15:37:44] [1] FATAL root %% - start of app [src/main.cpp:36]
Dec 12 16:37:46 solewer WARN: [ID 702911 local0.crit] [12 12 2011 15:37:46] [1] FATAL root %% - finish of app [src/main.cpp:41]

A nice feature to report hard errors to a central logging facility with nearly
no effort.

compiling log4cplus with SunStudio 12.3 beta on OpenIndiana

2011-12-12T15:46:00.000+01:00

I experimented with log4cplus on Windows and MinGW, which is an
exquisite suite for compiling on Windows.

Today I attempted to compile the library on my OpenIndiana 151 box.

The compilation with the GCC 4.6.1 was no problem.
The next attempt was to compile with the SunStudio Compiler 12.3 beta, which
is now called Oracle Solaris Studio ( found the former name better, because
the compiler suite is also available for linux which has nothing to do with
solaris ).

The compilation by configure does not continue without problems.
First the PTHREAD was not found -> changed path, so that SUNWspro
was found first and set CC=cc & CXX=CC
Configure was not a problem now, but compiling or to be exact the linking
didn't succeed.

CC: Warning: -xarch=amd64 is deprecated, use -m64 to create 64-bit programs
CC: -library=Cstd cannot be used with -library=stlport4

Therefore I had to manually edit libtool and remove the option -library=Cstd
and also the linking was fine.

Compiled for 32 and 64 bit.

log4cplus -> dynamic configuration of log during runtime

2011-12-11T17:10:00.001+01:00

Today I testetd around with the log4cplus documented here.

I used a simple configuration by properties like:

log4cplus::PropertyConfigurator::doConfigure( logFileName );

but now I found another possible config by a thread watcher:

log4cplus::ConfigureAndWatchThread watcher( logFileName.c_str(), 5*1000 );

This one checks every 5 seconds the configuration file and updateds the logging
framework.

So by changing the properties, the logging of the running process can be configured.
I've teste this with MinGW on Windows.

log4cplus on Windows with MinGW

2011-12-11T16:10:00.001+01:00

Today I tried the log4cplus Framework. Log4cplus is under the Apache License V2.0.
This one should be more usefull for commercial projects. The last time I wrote about log4cpp.
This framework is under the LGPL licencse. I've found somewhere, that it should be possible to
use in commercial products, but with some restrictions, about only dynamically linked, hint where to
get the source, allowing the customer to change the lib, ....

I've not red the Apache License in detail, because for me it's just an attempt to use it for some testing
and learning a little C++.

log4cplus can be configured and built simple be an MinGW environment on Windows.

For running an app, the DLL of log4cplus must be found on the PATH.

A property file can look like:

log4cplus.rootCategory=ERROR, rootAppender
log4cplus.logger.Names=DEBUG, A1, A2

log4cplus.additivity.Names=false

log4cplus.appender.rootAppender=log4cplus::ConsoleAppender
log4cplus.appender.rootAppender.layout=log4cplus::PatternLayout
log4cplus.appender.rootAppender.layout.ConversionPattern=%d{%m/%d/%y %H:%M:%S} [%t] %-5p %c{2} %%%x%% - %m [%l]%n

log4cplus.appender.A1=log4cplus::RollingFileAppender
log4cplus.appender.A1.File=A1.log
log4cplus.appender.A1.MaxFileSize=1MB
log4cplus.appender.A1.MaxBackupIndex=2
log4cplus.appender.A1.layout=log4cplus::SimpleLayout
log4cplus.appender.A1.layout.ContextPrinting=enabled
log4cplus.appender.A1.layout.DateFormat=ISO8601

log4cplus.appender.A2=log4cplus::ConsoleAppender
log4cplus.appender.A2.layout=log4cplus::PatternLayout
log4cplus.appender.A2.layout.ConversionPattern=[%d{%d %m %Y %H:%M:%S} [%t] %-5p %c{2} %%%x%% - %m [%l]%n

The main module :

#include <log4cplus/configurator.h>

...

    std::string logFileName = "log4cplus.prop";
    log4cplus::PropertyConfigurator::doConfigure( logFileName );    
    cout << "Hello World" << endl;

Initializing the logger:

    Names::Names() :firstName(""), lastName(""), log(log4cplus::Logger::getInstance("Names"))
    {
        LOG4CPLUS_INFO( log, "Default Constructor" );
    }

Using in a complex format:

        LOG4CPLUS_INFO( log, "FirstName " <<
                    this->firstName
                                         << "; LastName "
                    << this->lastName );

Example of output:

[11 12 2011 15:50:41] [4624] INFO Names %% - FirstName a; LastName b [Names.cpp:46]
[11 12 2011 15:50:41] [4624] INFO Names %% - Combined name <a b> [Names.cpp:54]

Update: A follow up describes how to allow configuration changes of the logging for a running
process.

log4cpp Usage

2011-12-08T14:55:00.001+01:00

The last time I compiled log4cpp on Windows with MinGW.

Now I got the following to work:
Names.hpp

#ifndef NAMES_HPP
#define   NAMES_HPP

#include <log4cpp/Category.hh>

namespace names {


   class Names
   {
   public:
        Names();
        Names(const Names& orig);
        virtual ~Names();
        void getNames();
        std::string myNames();
   private:
        std::string firstName;
        std::string lastName;
        log4cpp::Category& logger;

};
}
#endif /* NAMES_HPP */

using in Names.cpp as:

Names::Names(....) :logger(log4cpp::Category::getInstance("Names")
{
}

string Names::myNames()
    {
        std::stringstream combinedName;

        combinedName << this->firstName << " " << this->lastName;
        logger.info("Combined name <%s>", combinedName.str().c_str());
        return combinedName.str();
    }

and in the Main-Method:

int main(int argc, char** argv)
{
    std::string logFileName = "log4cpp.prop";
    try
    {
        log4cpp::PropertyConfigurator::configure(logFileName );
    }
    catch( log4cpp::ConfigureFailure e)
    {
        cerr << "Log4cpp Error: " << e.what() << endl;
    }

where the property configuration is:

log4j.rootCategory=DEBUG, rootAppender
log4j.category.Names=DEBUG, A2
log4j.additivity.Names=false

log4j.category.sub2=INFO
log4j.category.sub1.sub2=ERROR, A2

log4j.appender.rootAppender=org.apache.log4j.ConsoleAppender
log4j.appender.rootAppender.layout=org.apache.log4j.BasicLayout

log4j.appender.A1=org.apache.log4j.FileAppender
log4j.appender.A1.fileName=A1.log
log4j.appender.A1.layout=org.apache.log4j.BasicLayout

log4j.appender.A2=org.apache.log4j.ConsoleAppender
log4j.appender.A2.layout=org.apache.log4j.PatternLayout
log4j.appender.A2.layout.ConversionPattern=[%d{ISO8601}] [%c/%p] %m %n

and a sample output looks like:

[2011-12-08 14:55:00,182] [Names/INFO] Default Constructor
[2011-12-08 15:14:10,542] [Names/INFO] FirstName a; LastName b

For linking with MinGW, I had to add the following lib -lws2_32

log4cpp compile on Windows with MinGW and Usage in NetBeans 7.1RC1

2011-12-05T09:12:00.001+01:00

Just for testting, I attempted to use NetBeans 7.1RC1 on Windows with MinGW-Compiler.

I installed to MinGW installer, which was getting everything from the Web.

In NetBeans the Compiler must be configured by adding the MinGw/bin Path as base directroy.

Creating a new project still fails, because I had also to add
C:\MinGW\msys\1.0\bin to the PATH envrionment to find mkdir, which is used in
the makefiles of a C++ Project in NetBeans, otherwise the build fails.

log4cpp can be downloaded starting from here.

Starting a bash and using configure --prefix=/c/sdk/log4cpp/1.0 started the configure and
afterwards a make attempted to compile the source, but there are some problems.

I found the solution by a nice guy Zih Jun on his blog.

After some modifications in the log4cpp source, it could be compiled, although there was
a problem with dirname in the configure step.

After doing a make install all necessary things were copied.

In the Project-Options of NetBeans, I added the include path.

But the code completion just showed some files.

The reason therefore ist, that NetBeans doesn't recognise header files with the extension .hh which
log4cpp is using.

Adding hh to the extensions showed all headers.

Now the real learing starts, how to use log4cpp.

UrukDroid 1.6RC2 upgrade to 1.6RC4

2011-11-30T09:40:00.001+01:00

This week a 1.6RC3 update was released by sauron.
Yesterday I started a backup and an upgrade and found, that also
a 1.6RC4 is available.

Did the upgrade on the Archos 101 and had no problems.

On The RC2 there were some hangs but not very often.

See this forum for infos.

Running pdf_presenter_console on OpenIndiana

2011-11-17T12:30:00.001+01:00

On this blog, I have recognized the PdfPresenterConsole.

To compile it, there are some prerequsites. I had to manually compile the
vala-Language.

Evince already brought in the libpoppler and the remaining parts were
already on my laptop.

Doing the compilation issued an error from cmake about an unsupported
command unset.

Because I compiled a new vala compiler, the version must be suitable.
To circumvent this compile error, I commented in
Pdf-Presenter-Console-2.0/cmake/Vala_CMake/vala/ValaVersion.cmake

unset(version_accepted)

out.
After this the compilation could be done without a problem and
the presenter could be started without a problem.

It displays on the internal and external Screen, shows the remaining
presentation time and next slide.

VPN to Cisco via vpnc on OpenIndiana

2011-11-13T10:30:00.001+01:00

On my laptop, I have OpenIndiana 151a installed.
The sfe repository has a vpnc package, which I installed on my system.

I configured a site config file in /etc/vpnc/, the VPN tunnel seemd to be
established, but all changing of routing issued errors about
"network not reachable", although the IP's DNS-Server IP and so on
are retrived via the VPN tunnel.

After a message to the openindiana-discussion list I got some response
and a hint from Sriram Narayanan where he has a VPN tunnel.

On his supplied links:

http://my2ndhead.blogspot.com/2008/11/connecting-opensolaris-to-cisco-vpn.html
http://atechnicaljourney.blogspot.com/2008/10/howto-get-tuntap-compile-for-x64.html

I found the redirect to the VPNC scripts.
From there, I downloaded the snapshot from 2011-09-04 and replaced
the vpnc-script in /etc/vpnc with the one from the snapshot.

After this I restarted vpnc and the tunnel got established and routing could
be added.

Now I can directly communicate with the other side of the tunnel and
issue a mercurial pull and so on.

The tunnel is up now and even DNS resolving works.

jEdit attempt to reopen a file on a networkdrive not available

2011-11-13T09:25:00.001+01:00

jEdit ist a nice java editor supporting to open and edit files via a SFTP
server.
I think it's the FTP plugin which allows to do this.

But now, the last time I had a file open from a SFTP site and now this is not
available and the editor still attempts to get a connection to this site.
Opening a local directory is always denied because of an opening of the remote site.

There is a start option for jedit to do not open recent files

jedit -noremote

Now jEdit starts without trying to open remove files.

Thunderbird 8.0 & Firefox 8.0 on OpenIndiana

2011-11-09T14:04:00.000+01:00

Today I installed Thunderbird 8.0 & Firefox 8.0 on my OpenIndiana 151a.

So far everything is working. For thunderbird the lightning 1.0 must be installed.

The other plugins so far work, when the max version in install.rdf is increased
excpet the engimail. But engimail also didn't work in 7.0. Only in the 6.0 version with the engimail 1.3.2 from here.

I already attempted to compile engimail, but therefore I need a thunderbird
compilation environment, but I do not get thunderbird to compile in
OpenIndiana.

rinetd on AIX can not start

2011-11-09T13:02:00.000+01:00

if (!allowRules && allowRulesTotal > 0 ) {
goto lowMemory;
}
I've tried to compile rinetd on AIX 6.1 with gcc.

For the compilation the local getopt.c must be compiled and the header file
in rinetd.c must be #include "getopt.h" and all object files including getopt.o must be linked.

But starting the rinted issued the following error:

rinetd: not enough memory to start rinetd

After some looking around, in the source code, I found the problem.

I do not have any allow or deny rules in the rinetd.conf and therefore
the malloc resulted in a NULL Pointer which branched to the lowMemory mark.

old:

if (!allowRules) {
goto lowMemory;
}

new:

if (!allowRules && allowRulesTotal > 0 ) {
goto lowMemory;
}

old:

if (!denyRules) {
goto lowMemory;
}

new:

if (!denyRules && denyRulesTotal > 0) {
goto lowMemory;
}

Interesting is, why this was not a problem on e.g.: Solaris?

Updated VirtualBox to Versoin 4.1.6 from 4.1.4

2011-11-04T19:43:00.003+01:00

Today VirtualBox 4.1.6 is realeased. Here is the changelog.

Installation was done without problems. The start of my Win XP instance and
the update of the Guest Additions were no problems.

The instance is running.

The VirtualBox Manager can only be started by setting LD_NODIRECT=1, but the Preferences Dialog causes a core dump.

The extension pack can be installed via the commandline:

[ewer@solewer:~/tmp]$/opt/VirtualBox/VBoxManage extpack install --replace ../Downloads/Oracle_VM_VirtualBox_Extension_Pack-4.1.6-74713.vbox-extpack
0%...10%...20%...30%...40%...50%...60%...70%...80%...90%...100%
Successfully installed "Oracle VM VirtualBox Extension Pack".

The Settings of the VM can be viewed.

Archos Tablet Update to UrukDroid 1.6RC2

2011-11-03T08:22:00.000+01:00

Yesterday Sauron published RC2 of UrukDroid 1.6. The Update can be downloaded from http://sauron.pourix.com/UrukDroid/UrukDroid-1.6/.

The Update was smooth without problems. Only the GMail account has to
be inserted again.

The backup menu of UrukDroid is reachable by booting the Archos and
after disappearing of the Archos logo the volume key has to be pressed.
( But that I have to verify, because yesterday I did just got into the backup for
one time to save root & data filesystem. )

Upgraded VirtualBox to just release 4.1.4 on OpenIndiana 151a

2011-10-04T21:53:00.002+02:00

Just saw, that VirtualBox 1.4.1 is released.

Downloaded did a pkgrm SUNWvbox and a pkgadd.

VirtualBox started without problem, by using my startscript setting LD_NODIRECT=1.

Have to test, if this version works without the workaround.

Getting cppunit 1.12.1 to compile on OpenIndiana 151a

2011-10-04T10:05:00.004+02:00

The cppunit 1.12.1 has some problems to be compiled on Solaris or maybee even *nix with gcc.

First configure fails, regardless if sunstudio or gcc is used as compiler platform.
dirname has to many options

creating include/cppunit/config-auto.h - prefix CPPUNIT for config/config.h defines
dirname: extra operand `/ausl'
Try `dirname --help' for more information.
/* automatically generated config/ contrib/ doc/ examples/ include/ lib

The solution was found here. Replace all

/* automatically generated */

with

$ac_prefix_conf_INP

in the configure script.
After this the configure worked.

But compiling cppunit was failing

g++ -DHAVE_CONFIG_H -I. -I../../config -I../../include -I../../include -g -O2 -MT TestAssert.lo -MD -MP -MF .deps/TestAssert.Tpo -c TestAssert.cpp -fPIC -DPIC -o .libs/TestAssert.o
In file included from TestAssert.cpp:2:0:
../../include/cppunit/portability/FloatingPoint.h: In function ‘int CppUnit::floatingPointIsFinite(double)’:
../../include/cppunit/portability/FloatingPoint.h:43:21: error: ‘finite’ was not declared in this scope
The solutions was in a forum entry for cppunit.

--- cppunit-1.12.1/include/cppunit/portability/FloatingPoint.h
+++ cppunit-1.12.1/include/cppunit/portability/FloatingPoint.h

@@ -3,6 +3,11 @@
#include <cppunit/Portability.h>
#include <math.h>
+
+#if defined(__sun) && !defined(CPPUNIT_HAVE_ISFINITE) && defined(CPPUNIT_HAVE_FINITE)
+#include <ieeefp.h>
+ // <math.h> is still needed for usage of fabs in TestAssert.cpp
+#endif
CPPUNIT_NS_BEGIN

After this the compilation works.

I used the compiler commands from here to compile with the SunStudio compiler.
I also had to insert the math.h after ieeefp.h because of unreferenced fabs
error.

How to change passphrase of gpg keyring

2011-10-03T14:08:00.000+02:00

On my OpenIndiana I have gpg2 installed.
The keyring is from an older installation, with an old passphrase.

To change the passphrase list your keys with

gpg2 -k

after this you can edit the key with

gpg --edit-key <your key>
>passwd

on my side, a GTK Dialog opens to request the old and the new passphrase.

seahorse is a GUI app to do the same, available in the OpenIndiana repo

Thunderbird 7.0 and engimail on OpenIndiana does not work

2011-09-29T11:39:00.000+02:00

After the release of Thunderbird 7.0, I installed the version on my
OpenIndiana installation, but engimail fails to work, although the plugin
starts.

The debug shows the following error:

2011-09-29 11:36:06.561 [ERROR] enigmail.js: Enigmail.initialize: Error - Enigmime Service not available
2011-09-29 11:36:06.561 [DEBUG] enigmail.js: Enigmail.initialize: exception=TypeError: Components.classes[NS_IPCSERVICE_CONTRACTID] is undefined

The screenshot is:

Switching back to thunderbird 6.0 lets engimail work.

smime signed email on Android

2011-09-26T09:58:00.000+02:00

A colleague started to sign the emails with smime. On my Android Tablet
I'm using K9 mail for mail reading, but the signed e-mails can not be viewed.
To see this emails also, I added an app named Djigzo from the Android market.
This can also "decrypt" the email and it can be seen on the Android device.

VirtualBox in OpenIndiana suddenly doesn't start -> VBoxSVC core dumps

2011-09-16T10:11:00.000+02:00

I upgraded to the latest version and switched also the sfe repository to

sfe origin online http://pkg.openindiana.org/sfe/
sfe-encumbered origin online http://pkg.openindiana.org/sfe-encumbered/

After some digging ( execsnoop of DTraceToolkit ) I saw, that
VirtualBox starts VBoxSVC.

I started VBoxSVC manually and after launching VirtualBox the VBoxSVC process terminates.

Here is the pstack of the core showing the relevant thread:

----------------- lwp# 3 / thread# 3 --------------------
00000000003c9cc5 ???????? ()
fffffd7fff1febc5 _SUNW_Unwind_RaiseException () + 55
fffffd7ffdda3b69 __cxa_throw () + 59
fffffd7ff7ae2c99 _ZN3xml4FileC1ENS0_4ModeEPKcb () + 2c9
fffffd7ff7ae8531 _ZN3xml13XmlFileParser4readERK9RTCStringRNS_8DocumentE () + b1
00000000006acade _ZN8settings14ConfigFileBaseC2EPKN3com7Utf8StrE () + 13e
00000000006bcdc5 _ZN8settings14MainConfigFileC1EPKN3com7Utf8StrE () + 35
00000000006954bd _ZN10VirtualBox4initEv () + 2dd
0000000000696a49 _ZN10VirtualBox14FinalConstructEv () + 9
00000000006efe9d _Z21VirtualBoxConstructorP11nsISupportsRK4nsIDPPv () + 12d
fffffd7ff6e963ce _ZN16nsGenericFactory14CreateInstanceEP11nsISupportsRK4nsIDPPv () + 24
fffffd7ff6e57373 _ZN22nsComponentManagerImpl14CreateInstanceERK4nsIDP11nsISupportsS2_PPv () + 175
fffffd7ff5d9a9e1 _ZN18ipcDConnectService7OnSetupEjPK13DConnectSetupj () + af
fffffd7ff5d9a917 _ZN18ipcDConnectService17OnIncomingRequestEjPK10DConnectOpj () + 2d
fffffd7ff5d99071 _ZN14DConnectWorker3RunEv () + 211
fffffd7ff6e638de _ZN8nsThread4MainEPv () + 2e
fffffd7ff6e9487a _pt_root () + 90
fffffd7ff6e9493f _pt_iprt_root () + c
fffffd7ff7b1374c rtThreadMain () + 2c
fffffd7ff7b60321 _Z18rtThreadNativeMainPv () + 81
fffffd7fff1f6e44 _thrp_setup () + bc
fffffd7fff1f7110 _lwp_start ()

The function where the Exception is thrown is:

[ewer@solewer:~]$echo "_ZN3xml4FileC1ENS0_4ModeEPKcb" | c++filt
xml::File::File(xml::File::Mode, char const*, bool)

A truss showed that VirtualBox.xml in $HOME/.VirtualBox is attempted to open.

Before I removed the whole .VirtualBox directory, because of starting problems. Also the backup from zfs snapshot didn't solve the problem.

I found the hint for solving the problem in a Bug for VirtualBox.

Setting the Environment LD_NODIRECT=1 sovled the problem and let
VirtualBox start.

After the first start some time it also starts without the LD_NODIRECT Variable, but anytime later it fails again.

Getting engimal in thunderbird on OpenIndiana to work

2011-09-15T10:11:00.001+02:00

Here I've given the link to the actual engimail -plugin for thunderbird 6,
to allow signing/encrypting of e-mails.

The installation of the plugin had no problems, but executing gpg was
a problem.

I installed gpg2 from the sfe repository.

But execution of gpg2 didn't find pinentry -> The executeables are in /usr/lib/.

So I linked /usr/lib/pinentry-gtk-2 and /usr/lib/pinentry-curses to /usr/bin
and signing with engimail works now.

Engimail 1.3 for Solaris x86 & Thunderbird 6.0

2011-09-14T17:54:00.000+02:00

I have tried to compile thunderbird 6 from source, so that I can also compile
engimail.

But I failed for thundebird, because in same sources, there config said, that
the platform is not supported.

Today I found, that somebody else had more luck and compiled engimail 1.3.2

Find it here.

Now I have to configure everyhting, to get it up and running.

VirtualBox running in OpenIndiana using a vnic routed with Crossbow

2011-09-14T12:45:00.000+02:00

OpenIndiana supports by the Crossbow-Project virtual Networkinterfaces and
a virtual Switch.

Now I've changed the configuration of the VirtualBox Instance to use such a vnic
connected to a etherswitch.

ewer@solewer:~$ dladm show-etherstub
LINK
vswitch1
ewer@solewer:~$ dladm show-vnic
LINK         OVER         SPEED MACADDRESS        MACADDRTYPE         VID
vnic1        vswitch1     0      2:8:20:55:9e:bf   random              0
vnicvsop0    vswitch1     0      2:8:20:7b:84:43   random              0
vnicwinxp2   vswitch1     0      2:8:20:48:d2:6d   random              0

The vnic1 is used in the global zone of the machine, where routing is activated

ewer@solewer:~$ routeadm
              Configuration   Current              Current
                     Option   Configuration        System State
---------------------------------------------------------------
               IPv4 routing   enabled              enabled
               IPv6 routing   disabled             disabled
            IPv4 forwarding   enabled              enabled
            IPv6 forwarding   disabled             disabled

           Routing services   "route:default ripng:default"

Routing daemons:

                      STATE   FMRI
                   disabled   svc:/network/routing/ripng:default
                   disabled   svc:/network/routing/legacy-routing:ipv4
                   disabled   svc:/network/routing/legacy-routing:ipv6
                   disabled   svc:/network/routing/rdisc:default
                     online   svc:/network/routing/route:default
                     online   svc:/network/routing/ndp:default

For the vnic1 a static address object is created, to have the interface always
up and running:

ewer@solewer:~$ ipadm show-addr
ADDROBJ           TYPE     STATE        ADDR
lo0/v4            static   ok           127.0.0.1/8
vnic1/v4          static   ok           192.168.1.1/24
e1000g0/_a        static   ok           10.0.50.58/8
lo0/v6            static   ok           ::1/128

In VirtualBox the Interace is selected and operated in Bridged Mode.
The important thing is to set the MAC-Address to the same value as the
interface itself has.

ewer@solewer:~$ dladm show-vnic vnicwinxp2
LINK OVER SPEED MACADDRESS MACADDRTYPE VID
vnicwinxp2 vswitch1 0 2:8:20:48:d2:6d random 0

oterhwise the network connection will not work. Got the hint from here.

Additionally I have ipnat active, to map the outgoing traffic to the server interface ip address.

ewer@solewer:~$ cat /etc/ipf/ipnat.conf
map e1000g0 192.168.1.0/24 -> 0/32

ewer@solewer:~$ svcs ipfilter
STATE STIME FMRI
online 9:08:16 svc:/network/ipfilter:default

Update OpenIndiana on 12.09.2011

2011-09-12T21:01:00.000+02:00

The OpenIndiana pkg repository got two frequently updates.

Both didn't have any problem. They created each a new bootenvironment

Here the log from the first update:

root@solewer:~# pkg install package/pkg
                Packages to update:     1
           Create boot environment:    No
DOWNLOAD                                  PKGS       FILES    XFER (MB)
Completed                                  1/1     124/124      0.7/0.7

PHASE                                        ACTIONS
Update Phase                                 240/240

PHASE                                          ITEMS
Package State Update Phase                       2/2
Package Cache Update Phase                       1/1
Image State Update Phase                         2/2

PHASE                                          ITEMS
Reading Existing Index                           8/8
Indexing Packages                                1/1
root@solewer:~# pkg update
                Packages to update: 1145
           Create boot environment:   Yes
DOWNLOAD                                  PKGS       FILES    XFER (MB)
Completed                              1145/1145 19518/19518 601.2/601.2

PHASE                                        ACTIONS
Removal Phase                              6183/6183
Install Phase                              7285/7285
Update Phase                             30378/30378

PHASE                                          ITEMS
Package State Update Phase                 2290/2290
Package Cache Update Phase                 1145/1145
Image State Update Phase                         2/2

PHASE                                          ITEMS
Reading Existing Index                           8/8
Indexing Packages                          1145/1145
Indexing Packages                          1145/1145
Optimizing Index...

PHASE                                          ITEMS
Indexing Packages                          1319/1319

A clone of OI151-6 exists and has been updated and activated.
On the next boot the Boot Environment OI151-7 will be mounted on '/'.
Reboot when ready to switch to this updated BE.

---------------------------------------------------------------------------
NOTE: Please review release notes posted at:

http://wiki.openindiana.org/oi/oi_151a+Release+Notes
---------------------------------------------------------------------------

The qemu-kvm still does not work for me:

[ewer@solewer:~]$pfexec qemu-kvm -hda /virtual/qemu.img -cdrom /virtual/debian.0.2.1-i386-netinst.iso -boot d -localtime -no-hpet
kvm: unhandled exit 80000021
kvm_run returned -22

KVM / QEMU on OpenIndiana 151a fails

2011-09-06T17:12:00.000+02:00

I attempted to use KVM from OpenIndiana running Build 151a.

After installing the packages, the kvm module is missing, even after reboot.
I had to modload kvm from the directory /usr/kernel/drv/amd64

First it failed, because the /dev/kvm device was missing. I did an
devfsadm, which seemd to create the device.

But starting qemu-kvm still fails with a looping

root@solewer:/usr/kernel/drv/amd64# qemu-kvm -cdrom /virtual/xpimg.iso -boot d /virtual/aa.img 2>&1 | more
KVM internal error. Suberror: 2
extra data[0]: 80000008
extra data[1]: 80000b08

Even just starting without options fails:

root@solewer:/usr/kernel/drv/amd64# qemu-kvm -monitor stdio
QEMU 0.14.1 monitor - type 'help' for more information
(qemu) kvm: unhandled exit 80000021
kvm_run returned -22

(qemu) info kvm
kvm support: enabled

Update:
After rebooting the day after, VirtualBox failed to start and locks the
whole computer.
After modunload -i <number of kvm module>, VirtualBox starts again.

mucommander can not sftp to OpenIndiana

2011-08-31T12:33:00.000+02:00

The muCommander, a file browser like Norton Commander, has the option
to connect to a remote server via sftp.

The connection to OpenIndiana fails. The /var/adm/messages log shows the following:

Aug 31 11:02:50 solewer sshd[7303]: [ID 800047 auth.crit] fatal: Client and server could not agree on a common cipher: client "blowfish-cbc,3des-cbc", server "aes128-ctr,aes192-ctr,aes256-ctr,arcfour128,arcfour256,arcfour". The server cipher list can be controlled using the "Ciphers" option, see sshd_config(4) for more information.

After adding the Ciphers listed in the man page for sshd_config containing
blowfish-cbc,3des-cbc
sshd_config:

Ciphers aes128-ctr,aes192-ctr,aes256-ctr,aes128-cbc,aes192-cbc,aes256-cbc,arcfour,arcfour128,arcfour256,3des-cbc,blowfish-cbc

and restarting the ssh service

svcadm restart ssh

muCommander can connect via sftp.

cvs2hg converted repository with branches and hg convert with filemap destroys data

2011-08-24T15:42:00.000+02:00

I had to convert a CVS repository to Mercurial. The CVS repository had a lot
of branches. The conversion was done, as described in my former post about converting a CVS to HG.

After the conversion, there was a root directory in the HG-Repo, which should
be removed.

hg convert --filemap map.txt <source> <target>

would do the magic.

The content of the filemap was:

rename <rootDir> .

But the resulting HG repository was missing some directories and files.
So this was a not usable for me.
I found the following issue discussing the problem.

The other way to convert was to convert from CVS to SubVersion and
use hg to convert from SubVersion to Mercurial

ATTENTION: for this you need subversion bindings for Python

cvs2svn --config=svn.cfg

The changes in the svn.cfg from the example were

ctx.output_option = NewRepositoryOutputOption(

to set the SVN Repo

change all the decorders to latin1

and set

run_options.add_project

the path to the CVS Repository.

After converting to SubVersion, I deployed the Repo to the Apache/WebDav
installation and started the conversion to HG by

hg convert -A authormap http://<SubVersionServer/Repo> <MercurialRepo>

and after this a run with

hg convert --filemap map.txt <MercurialRepo> <strippedRepo>

and the files were in the new repo with the removed root directory.

converting a CVS repository with branches to Mercurial ( hg )

2011-08-23T14:45:00.002+02:00

I had to convert a old CVS repository with a lot of branches to Mercurial.

The first attempt was to use the convert extension from Mercurial to do
a direct conversion from cvs to hg.

The conversion with

CVSROOT=$HOME/CVSDIR
export CVSROOT
cvs co

hg convert -A $HOME/authormap <cvsworkdir> <newHgRepo>

The conversion went through with a OutOfMemory during the run,
but hg convert can continue where it left the job.

But this new HG-Repository had wrong branch data inside.
After searching for a solution, I found out that the hints all said to
convert first to subversion and afterwards to mercurial.

Here is the home for repository conversion@Mercurial

I started by using cvs2hg which uses internally the cvs2svn code base.

After cloning the cvs2hg repo with

hg clone http://hg.gerg.ca/cvs2svn

and adjusting the config example I started to do the first run:

./cvs2hg --options=./cvs2hg-example.options

My settings changed in the example config where

to set the decoders to latin1,
set the author_transforms dictionary
set ctx.output_option for the name of the new hg repository.
set in run_otpions.set_project the path to the CVS repository to convert.

The CVS-Repository was local on my machine instead of a network drive,
because hg convert also likes to have the repo locally.

The first run failed, because on my OpenIndiana there was a wrong dbm implementation in python for the module anydbm.

I downloaded the gdbm 1.8.3, compiled and installed in the default
path to /usr/local.
After this I had to download Python 2.7 and compile it locally.

First edit Module/Setup in the python source tree:

gdbm gdbmmodule.c -I/usr/local/include -L/usr/local/lib -lgdbm

and some other modules inside the tree.
I had to set LD_LIBRARY_PATH to /usr/local/lib and /usr/sfw/lib for
gdbm and the gcc libs.

./configure --with-dbmliborder=gdbm --prefix <my installation dir>

After compiling and installing the new python,
a restart of the cvs2hg went through the conversion,l
and created a new HG repository.

UrukDroid CPU Governor - battery life

2011-08-22T10:18:00.000+02:00

After Updating my Archos to UrukDroid 1.5.1, the setting for the
cpugovernor as described here was "lost".

So I updated the /etc/uruk.conf/cpugovernor again t switch to the convservative governor.

During the first days of using UrukDroid, I thought, that the battery life was shortened.

After some time I recognized that sending the device to sleep does not always
work.
The green led beside the power button didn't always turn off. So I'm checking
the device now, if the led goes off and now the batter life is better.

New Update on OpenIndiana from 19.08.2011

2011-08-21T15:21:00.003+02:00

There are new updates on http://pkg.openindiana.org/dev-il available.

The update was smooth without problems after the reboot:

root@solewer:~# pkg image-update
                Packages to update:   356
           Create boot environment:   Yes
DOWNLOAD                                  PKGS       FILES    XFER (MB)
Completed                              356/356   3330/3330 168.7/168.7

PHASE                                        ACTIONS
Removal Phase                                986/986
Install Phase                                983/983
Update Phase                               5210/5210

PHASE                                          ITEMS
Package State Update Phase                   712/712
Package Cache Update Phase                   356/356
Image State Update Phase                         2/2

PHASE                                          ITEMS
Reading Existing Index                           8/8
Indexing Packages                            356/356
Indexing Packages                            356/356
Optimizing Index...

PHASE                                          ITEMS
Indexing Packages                          1288/1288

A clone of OI151-5 exists and has been updated and activated.
On the next boot the Boot Environment OI151-6 will be mounted on '/'.
Reboot when ready to switch to this updated BE.

---------------------------------------------------------------------------
NOTE: Please review release notes posted at:

http://docs.sun.com/doc/821-1479
---------------------------------------------------------------------------

Thunderbird 6.0 is available

2011-08-17T18:51:00.001+02:00

Thunderbird 6.0 is available for download.

Works so far on OpenIndiana, but Lighting does not work ( 1.0b5) even when
updating the install.rdf to the new version, there are some errors.

Exception: [Exception... "Component returned failure code: 0x80570015 (NS_ERROR_XPC_CI_RETURNED_FAILURE) [nsIJSCID.createInstance]" nsresult: "0x80570015 (NS_ERROR_XPC_CI_RETURNED_FAILURE)" location: "JS frame :: resource://calendar/modules/calUtils.jsm -> file:///export/home/ewer/.thunderbird/hlimkdtv.default/extensions/%7Be2fda1a4-762b-4020-b5ad-a41df1933103%7D/calendar-js/calUtils.js :: createTodo :: line 67" data: no]
1: [file:///export/home/ewer/.thunderbird/hlimkdtv.default/extensions/%7Be2fda1a4-762b-4020-b5ad-a41df1933103%7D/components/calStorageCalendar.js:2364] cSC_logError
2: [file:///export/home/ewer/.thunderbird/hlimkdtv.default/extensions/%7Be2fda1a4-762b-4020-b5ad-a41df1933103%7D/components/calStorageCalendar.js:844] cSC_getItems_
3: [file:///export/home/ewer/.thunderbird/hlimkdtv.default/extensions/%7Be2fda1a4-762b-4020-b5ad-a41df1933103%7D/components/calStorageCalendar.js:647] null

This is a no go for me.

Update 19.08.2011:

There is an update for lightning 1.0b5
having a Thunderbird 6.0 compatible version at http://releases.mozilla.org/pub/mozilla.org/calendar/lightning/releases/1.0b5/contrib/solaris-x86/

With this update, the calendar plugin works again in Thunderbird 6.0

The quicktext extension has also an update from the developer ( http://extensions.hesslow.se/extension/4/Quicktext/ ) to let it work on
Thunderbird 6.0

Switched over to Thunderbird 6.0

VirtualBox 4.2.1 is released, but USB seems not to work in OpenIndiana

2011-08-16T18:28:00.000+02:00

VirtualBox 4.2.1 is release.
Installation on OpenIndiana Build 151 with Illumos repository works.
Windows XP starts, but USB-Devices seem not to work:

Also a modinfo shows, that the module is loaded

root@solewer:/opt/VirtualBox# modinfo | grep vbox
219 fffffffff8b49000 2e370 296   1 vboxdrv (VirtualBox HostDrv 4.1.2r73507)
259 fffffffff7921550    d48 297   1 vboxnet (VirtualBox NetAdp 4.1.2r73507)
261 fffffffff8d33000   7960 298   1 vboxflt (VirtualBox NetDrv 4.1.2r73507)
261 fffffffff8d33000   7960   -   1 vboxflt (VirtualBox NetMod 4.1.2r73507)
262 fffffffff8d3b000   4a08 299   1 vboxusbmon (VirtualBox USBMon 4.1.2r73507)
263 fffffffff8d40000   75d8 300   1 vboxusb (VirtualBox USB 4.1.2r73507)

Firefox 6.0 available for download.

2011-08-15T13:42:00.001+02:00

Firefox 6.0 is available for download.
Windows, Linux, Solaris see here.

SSH access to UrukDroid 1.5

2011-08-15T13:40:00.000+02:00

First activate via UrukConfig the SSH server.

After activating SSH, teh overview of services shows the IP adress of the device.

Use putty.exe on Windows to connect to the device.
Username is root
Password is UrukDroid ( found here )

Changing cpugovernor on UrukDroid

2011-08-15T13:35:00.004+02:00

The UrukConfig app allows to activate the CPU governor to activate the dynamic CPU
switching.

To save battery power, I changed in /etc/uruk.conf/cpugovernor

cpu_governor="conservative"
...
cpu_governor_modules="cpufreq_powersave cpufreq_conservative cpufreq_userspace"

cd /etc/uruk.d
./cpugovernor stats

shows the ticks in each frequency range.

During tests with 1.5B3 it showed a lot more ticks in 300 MHz instead of ticks in 1GHz mode.

The conservative mode only increased the CPU frequenzy for some amount and does not switch to
full mode.

UrukDroid 1.5 update from 1.5RC1 on Archos 10.1

2011-08-15T10:37:00.002+02:00

On http://sauron.pourix.com/UrukDroid/UrukDroid-1.5/ is a new update file for 1.5
I first had to update with UrukDroid-1.5-kernel.tbz2 by copy of the file to /data/UrukUpdate.
There was no autoamtic reboot, but looking into /var/lib/urukdroid/update.log, showed that the
kernel update was successful.
There ware 2 killed message in /system/xbin/uruk_upd: line 94

After this installation, the start showed uruk 1.5 as boot, but swithing to the rootfs still shows 1.5RC1.
Also putting UrukDroid-rootfs-update.tgz into /.upgrade and rebooting didn't remove the 1.5RC1 from swithing to rootfs

Since midday ( 15.08.2011 ) there is also a UrukDroid-1.5-upgrade.tbz2 upgrade file to be put to /data/UrukUpdate.

After putting it into the directory, an automatic reboot is initiated after the uprade es decompressed and
during the reboot the rootfs is upgraded and the 1.5RC1 for init of rootfs is 1.5 and not any RC1 more.

cat /etc/urukdroid-version

shows

UrukDroid 1.5 ( 15.08.2011)

UrukDroid 1.5 RC1 is available for download

2011-08-10T21:38:00.004+02:00

On my Archos 10.1 if've installed UrukDroid, to get root access and vpnc ( for VPN connection ) running.

Now there is the installation for 1.5 RC1 available.

I was allready on B3 and installed. But this was a mistake from my side. It really installed it and works
fine, but my installation with the apps was trashed.

But this was an opportunity to reformat.

The installation was via the SDE menu Recovery System -> Update Firmware and installed the firmware

Started via SDE the Developer Edition and copied the install-File.

This installation works without problems.

New UrukDroid Update 1.5B3 for Archos available

2011-08-05T20:53:00.000+02:00

There is a new Update for Archos 101 UrukDroid 1.5 available. This is B3.

The direct update by coping to /data/UrukUpdate doesn't work. This shows "wrong signature" in
the update.fail log -file in ( /var/lib/urukdroid I think so )

But on xda-developers there is described how to manually update the system.

0. Extract UrukDroid-1.5B3-upgrade.tbz2
1. Flash zImage and initramfs.cpio.gz through SDE
2. Enter USB MMC and copy UrukConfig.apk to /data/app and bootanimation.zip to /data/customization
3. Create .upgrade folder in / and put UrukDroid-rootfs-upgrade.tgz in it.
4. Reboot and the upgrade will start.

Upgrade the manual way worked for me.

Changed OS on Archos 101 to UrukDroid ( upgrade to 1.5B2 )

2011-07-31T11:04:00.003+02:00

Now I've changed to OS on my Archos 101 from the Archos firmware to UrukDroid.

First contact was an installation tutorial for 1.0.
Where also the files are listed for download.
The homepage of UrukDroid is at google.

To use UrukDroid, the Archos must be patched to use the SDE Firmware from Archos itself.

Following the installation instruction was no problem to install.

The next Versions of UrukDroid are not released yet, but I'd like to upgrade to the Beta-Version
of 1.5. The download files root is here.

My first fault was, to extract the archive and just copy oer the update-File, but you have to
copy the whole archive downloaded to the archos device.

The archive has to be put to /data/UrukUpdate on you're device. So you have to use a Fileexplorer or
terminal to move the file from /sdcard to /data/UrukUpdate.
After some time the device starts to update itself.

I recognized a slowdown of the device during the updated phase. When the update is finished, the
devie reboots itself.

The file /etc/urukdroid-version contains the actual version of the instaled system.

Also vpnc ( VPN Network establish tool ) works now with UrukDroid 1.5.
Version 1.0 was not able to establish a VPN Connection.

Updated my OpenIndiana Build 151 Installation

2011-07-26T15:23:00.002+02:00

Just saw that, the there was an update in PKG-Repository for Dev-Illumos at
25.07.2011.

So I started a pkg upate

root@solewer:~# pkg update

                Packages to update:   350
           Create boot environment:   Yes
DOWNLOAD                                  PKGS       FILES    XFER (MB)
Completed                              350/350   1522/1522    24.0/24.0

PHASE                                        ACTIONS
Removal Phase                                  99/99
Install Phase                                  80/80
Update Phase                               5414/5414

PHASE                                          ITEMS
Package State Update Phase                   700/700
Package Cache Update Phase                   350/350
Image State Update Phase                         2/2

A clone of OI151-2 exists and has been updated and activated.
On the next boot the Boot Environment OI151-3 will be mounted on '/'.
Reboot when ready to switch to this updated BE.

---------------------------------------------------------------------------
NOTE: Please review release notes posted at:

http://docs.sun.com/doc/821-1479
---------------------------------------------------------------------------

After the reboot everything seems to be fine so far.

C/C++ Feature in NetBeans to show the Callgraph of a function

2011-07-25T15:29:00.000+02:00

Found a nice feature in the C/C++ Plugin of Netbeans.

Select a function in the code, right click and select "Show Call Graph".
This opens a window and shows the callgraph depending on the selection you do.
All Functions called by this one or all other functions calling this one.

VirtualBox 4.1.0 is release

2011-07-20T09:58:00.001+02:00

The Version 4.1.0 of VirtualBox is released.

Download for Solaris x86/amd64 is here.

Installation and start of an XP instance in OpenIndiana 151 was not problem.
The installation still needed the saved boot of XP.

VirtualBox 4.0.12 ist released

2011-07-18T16:03:00.002+02:00

A little bit late, because of vacation, but Version 4.0.12 of VirtualBox is release.
Downloaded and installed on OpenIndiana without problems. The Virtual XP
instance is up and running with the updated Guest Additions.

Thunderbird 5.0 -> QuickText extensions does not work

2011-07-18T16:00:00.000+02:00

I installed Thunderbird 5.0 on my OpenIndiana Box from the reference page in opensolaris.org.

In former releases of Thunderbird, I used a lot the quicktext-extension.
This extension allows to define macros/templates of text, which can be
expanded after pressing a special key like TAB in my case.
But the extension does not work anymore with Thunderbird 5.0.
Also editing the install.rdf and increasing the version number does not solve
the problem.

But at the home of the developer I found a beta version 0.9.11.0 for download.
This extension also works in thunderbird 5.0 and all my templates are
here again.

Bugfix-Release von VirtualBox 4.0.10 -> installation on OpenIndiana 151 is ok

2011-06-28T16:54:00.001+02:00

VirtualBox released a Bugfix version 4.0.10 for Solaris . ( download )
Here is the changelog.

Uninstalling 4.0.8 and installing 4.0.10 on OpenIndiana Build 151 was done
without problems.

The XP installation was rebooted and the new GuestAdditions are installed.
No problems with the XP virtual machine here.

Christening Cake

2011-06-22T18:48:00.003+02:00

Installing Oracle 11g in an OpenIndiana Zone

2011-06-21T15:28:00.004+02:00

For the Installer of Oracle X11 is necessary.

Beside the installation of a zone and adding the xauth-Package to allow X11-Forwarding in ssh, the Oracle Installer requires the motif library libXm

To get the installer up and running just install motif in the zone:

pkg install motif
pkg install xauth
pkg install developer/object-file

For the GUI-Installer of oracle I set the LANG to de_AT.UTF-8 because of
the following error:

ERROR: Unable to convert from "UTF-8" to "646" for NLS!

I had to install locale/de-extra

After this, the runInstaller from Oracle database worked.

For detailed installation instructions look here.

X11 forwarding from an OpenIndiana Zone

2011-06-21T15:25:00.005+02:00

For testing, I installed a zone on OpenIndiana Build 151.
(zonecfg, zoneadm install ... )
Logging in via ssh works over a crossbow switch. /etc/ssh/sshd_config is configured
to allow X11 Forwarding, but a ssh -X call fails.

Enabling debugging of ssh with ssh -Xv showed, that xauth is missing in the zone.

pkg install xauth

solved this issued for me.

Compiling pycrypto 2.3 and paramiko 1.7.7.1 on OpenIndiana/Solaris

2011-06-03T20:30:00.002+02:00

For using SSH in python you have to get paramiko and pycrypto.

I'm using the packages on OpenIndiana 151.

For the compilation I used Sun Studio 12.2.

The only thing I had to do was setting PYCC_CC to the cc of the Sun Studio
12.2 Compiler package and change the include of gmp.h in src/_fastmath.c
to gmp/gmp.h in the pycrypto-Source.

After this change, the compilation went fine through.

>>> os.uname()
('SunOS', 'solewer', '5.11', 'oi_151', 'i86pc')
>>> import Crypto
>>> Crypto.__version__
'2.3'
>>> import paramiko
>>> paramiko.__version__
'1.7.7.1 (George)'

Help for paramiko is here.

Upgrade from OpenIndiana 148 to OpenIndiana 151

2011-05-28T08:30:00.001+02:00

Now I've upgraded my OpenIndiana Build 148 Installation to
OpenIndiana Build 151 on http://pkg.openindiana.org/dev-il, which is
an Illumos based version.

root@solewer:~# pkg image-update --be-name OI151 --require-new-be
                Packages to remove:    86
               Packages to install:    72
                Packages to update: 1053
           Create boot environment:   Yes
DOWNLOAD                                  PKGS       FILES    XFER (MB)
Completed                              1211/1211 298root@solewer:~# pkg image-update --be-name OI151 --require-new-be
                Packages to remove:    86
               Packages to install:    72
                Packages to update: 1053
           Create boot environment:   Yes
DOWNLOAD                                  PKGS       FILES    XFER (MB)
Completed                              1211/1211 29846/29846 815.0/815.0

PHASE                                        ACTIONS
Removal Phase                            19343/19343
Install Phase                            14309/14309
Update Phase                             43342/43342

PHASE                                          ITEMS
Package State Update Phase                 2264/2264
Package Cache Update Phase                 1139/1139
Image State Update Phase                         2/2

A clone of OI_148 exists and has been updated and activated.
On the next boot the Boot Environment OI151 will be mounted on '/'.
Reboot when ready to switch to this updated BE.
46/29846 815.0/815.0

PHASE                                        ACTIONS
Removal Phase                            19343/19343
Install Phase                            14309/14309
Update Phase                             43342/43342

PHASE                                          ITEMS
Package State Update Phase                 2264/2264
Package Cache Update Phase                 1139/1139
Image State Update Phase                         2/2

A clone of OI_148 exists and has been updated and activated.
On the next boot the Boot Environment OI151 will be mounted on '/'.
Reboot when ready to switch to this updated BE.

So far I do not see any problems.
Upgraded VirtualBox to V4.0.8 and my XP-Instance boots

Suspend to ram is ok, resume starts, but never finishes.
On Solaris 11 Express with Build 151 resume works.
The nvidia driver is also the beta driver listed on the OpenIndiana Website for Build 151.

There are no messages in /var/adm/messages.

How to show the related files to a Mercurial MQ patch

2011-05-12T14:31:00.000+02:00

The distributed version control system mercurial allows to use
MQ ( Mercurial queues ) to have patcsets collected, which can be
applied over a repository.

When using tortoisehg or the mercurial delivered tool hgtk ( hg view ) you
can see the related files belonging to a patchset.

The command qrefresh of mercurial allows to update a patchset.
But when you have different modified files, you don't want to have all
modifications in the patchset.
qrefresh offers to list the files on th run, but the problem is, that only those
listed files are in the patchset. You also have to add the files already in the
patchset.

The GUI-Tools list the files belonging to a changeset, but how to do it on the
command line?

The command status is for help here. Without any option, the command
list the statue of the working directory, but when you call it with the
option --change , you get the files for the changeset:

hg status --change qtip

This command lists the files belonging to the last patch applied, but instead
of qtip you can also take the Patchname.

Not able to start another process on AIX

2011-05-10T15:00:00.000+02:00

If you've problems to start new processes running under a single user on AIX,
it could be the operating system AIX which, blocks to start a new process.

The default number of processes per user is 128 on AIX:

smitty chgsys

You can reach the menu in smitty via

System Environments -> Change/Show Characteristics of Operating System

How to get new locales in a debian system.

2011-05-10T14:23:00.000+02:00

I have to port an application to a debian squeeze system.

The system was already installed by someone else, so I had to add
some additional packages:

It is a 64bit system, but I'd like to compile my sources also in 32 bit mode.

To be able to do this, I had to add the packages gcc-multilib and g++-multilib
to allow compilation with C/C++ in 32 Bit mode on the debian system.
The tool for installation I used was aptitude - a curses based installer.

Because the target platform is a german one, I'd also liked to have some
locales like de_DE.ISO8859-15, but the system just had

#locale -a
C
POSIX
de_AT.utf8

installed.

To add some more locales I found the following Debian-wiki entry for locale:
The command for me was:

dpkg-reconfigure locales

This allowed in a curses based dialog to select the locales to generate on the system.

Netbeans 7.0 and Python

2011-04-29T08:10:00.000+02:00

As it seems, from the last message on the mailinglist of the python plugin for Netbeans, the project seems to be shut down.

So I think I've to switch over to Eclipse with the pydev-Plugin for my
python editing.

Hopefully there will sometime be a comeback of python support in Netbeans.

VirtualBox 4.0.6 installed on Solaris 11 Express

2011-04-29T08:05:00.000+02:00

I've installed VirtualBox 4.0.6 from the download page on the Instance
of Solaris 11 Express. My virtualised XP instance runs without problems.

Only when the System is sent into suspend mode, the running instance must
be closed, because otherwise the system won't not wake up again.

Netbeans 7.0 released and available for download

2011-04-21T09:20:00.000+02:00

The new NetBeans 7.0 is available for download.

I've installed on Solaris 11 Express and Solaris 10 x86.

The Python Modules are currently missing in the plugincenter.

Python Module of the Week - blog

2011-04-20T09:28:00.000+02:00

After viewing a video from the Python Conference 2011 about

Hidden Treasures in the Standard Library

I got a valuable hint about the blog "Python module of the week"
written by the speaker Doug Hellmann.

I just viewed some entries, and they describe in a detailed way the
functions in the standard python library.

New Update of Firefox and thunderbird for Solaris/OpenSolaris

2011-03-07T10:28:00.000+01:00

There are new updates for Firefox ( 3.6.15, 3.5.17, 4.0b12 ) and thunderbird ( 3.1.9 )

Go here for the links.

How to insert Umlaute like ÜÖÄ with my Archos 10.1

2011-01-29T17:11:00.001+01:00

To insert Umlaute via the screen keyboard of my Archos 10.1 tablet, I have to press for alonger time on o a or u and the alternative characters are displayed.

VirtualBox 4.0.2 Update is available

2011-01-19T09:51:00.002+01:00

From the download-Site of VirtualBox the Update 4.0.2 is available.
The changelog is here.

Uninstalling the last 4.0.0 package on Solaris 11 Express and installing
the 4.0.2 Update was no problem. Also the VBoxAdditions intalled on the
Windows XP in VirtualBox.

Resume work on Solaris 11 Express and fails on OpenIndiana 148

2011-01-19T09:03:00.000+01:00

Last week I installed on OpenSolaris Build 134 Solaris 11 Express, beside
OpenIndiana 148.

In Solaris 11 Express the suspend/resume works on my laptop

# prtdiag -v
System Configuration: FUJITSU SIEMENS CELSIUS H270
BIOS Configuration: Phoenix Z118D4FR.113 - R081 - 4807 12/11/2009

But on OpenIndiana the resume fails. I think this depends on the driver for
the nvidia card.

Released VirtualBox 4.0.0 installed on OpenIndiana Build 148

2010-12-23T10:09:00.000+01:00

The Version 4.0.0 of VirtualBox is released as a Christmas present.

After downloading, the installation was fine and the installed Windows XP is
up and running in Seamless mode.

Additional to the package, there is a extension pack available now.

See here for a documentation. The extension pack provided by Oracle contains
support for some additional hardware - see the link above.

Upgraded from OpenIndiana 147 to OpenIndiana 148

2010-12-18T10:51:00.000+01:00

Today I upgraded from OpenIndiana 147 to OpenIndiana 148 without any
problems.

root@solewer:~# pkg image-update --be-name OI_148
                Packages to remove:    10
               Packages to install:    11
                Packages to update: 1095
           Create boot environment:   Yes
DOWNLOAD                                  PKGS       FILES    XFER (MB)
Completed                              1116/1116 22673/22673 780.6/780.6

PHASE                                        ACTIONS
Removal Phase                              6579/6579
Install Phase                              5186/5186
Update Phase                             36358/36358

PHASE                                          ITEMS
Package State Update Phase                 2211/2211
Package Cache Update Phase                 1105/1105
Image State Update Phase                         2/2

A clone of OI_147 exists and has been updated and activated.
On the next boot the Boot Environment OI_148 will be mounted on '/'.
Reboot when ready to switch to this updated BE.

[ewer@solewer:~]$uname -a
SunOS solewer 5.11 oi_148 i86pc i386 i86pc Solaris

The restart imported 4 new SMF-Services and delete the javadb-Service because of a missing manifest.

The suspend/resume still doesn't work.

nvidia seems to work on this laptop

nVidia Corporation G96M [Quadro FX 770M]

working with Mercurial behind a proxy

2010-12-14T16:23:00.000+01:00

Mercurial can use the http protocol for version management, but when you
are behind a proxy, you can simple set the configuration in your hgrc file
or give the configuration in the command line like this:

hg --config http_proxy.host=#IP#:#PORT# pull

VirtualBox 3.2.12 is installable on OpenIndiana 147 without any modification

2010-12-02T14:22:00.000+01:00

Today, I installed the VirtualBox 3.2.12. I remembered to have to
patch a file in the package to allow the installation on OpenIndiana.
So I transfered the stream package with pkgtrans and looked into the file,
but there are already changes in the package to recognize the OpenIndiana
distribution. So there is no change necessary anymore.

See here for the discussion at openIndiana.

SFTP withing jedit doesn't work to login in to SLES10 x 64

2010-11-26T10:38:00.000+01:00

I've a SLES10 x64, to which I'd like to connect with the SFTP Plugin in
jEdit to view remote files.

But the login fails with the following trace:

10:18:56 [jEdit I/O #2] [message] SftpLogger: Connection established
10:18:56 [jEdit I/O #2] [message] SftpLogger: Remote version string: SSH-1.99-OpenSSH_4.2
10:18:56 [jEdit I/O #2] [message] SftpLogger: Local version string: SSH-2.0-JSCH-0.1.42
10:18:56 [jEdit I/O #2] [message] SftpLogger: CheckCiphers: aes256-ctr,aes192-ctr,aes128-ctr,aes256-cbc,aes192-cbc,aes128-cbc,3des-ctr,arcfour,arcfour128,arcfour256
10:18:56 [jEdit I/O #2] [message] SftpLogger: aes256-ctr is not available.
10:18:56 [jEdit I/O #2] [message] SftpLogger: aes192-ctr is not available.
10:18:56 [jEdit I/O #2] [message] SftpLogger: aes256-cbc is not available.
10:18:56 [jEdit I/O #2] [message] SftpLogger: aes192-cbc is not available.
10:18:56 [jEdit I/O #2] [message] SftpLogger: arcfour256 is not available.
10:18:56 [jEdit I/O #2] [message] SftpLogger: SSH_MSG_KEXINIT sent
10:18:56 [jEdit I/O #2] [message] SftpLogger: SSH_MSG_KEXINIT received
10:18:56 [jEdit I/O #2] [message] SftpLogger: kex: server->client aes128-ctr hmac-md5 none
10:18:56 [jEdit I/O #2] [message] SftpLogger: kex: client->server aes128-ctr hmac-md5 none
10:18:56 [jEdit I/O #2] [message] SftpLogger: SSH_MSG_KEXDH_INIT sent
10:18:56 [jEdit I/O #2] [message] SftpLogger: expecting SSH_MSG_KEXDH_REPLY
10:18:56 [jEdit I/O #2] [message] SftpLogger: ssh_rsa_verify: signature true
10:18:56 [jEdit I/O #2] [message] SftpLogger: Host 'xxxxxxx' is known and mathces the RSA host key
10:18:56 [jEdit I/O #2] [message] SftpLogger: SSH_MSG_NEWKEYS sent
10:18:56 [jEdit I/O #2] [message] SftpLogger: SSH_MSG_NEWKEYS received
10:18:56 [jEdit I/O #2] [message] SftpLogger: SSH_MSG_SERVICE_REQUEST sent
10:18:56 [jEdit I/O #2] [message] SftpLogger: SSH_MSG_SERVICE_ACCEPT received
10:18:56 [jEdit I/O #2] [message] SftpLogger: Authentications that can continue: publickey,keyboard-interactive,password
10:18:56 [jEdit I/O #2] [message] SftpLogger: Next authentication method: publickey
10:18:56 [jEdit I/O #2] [message] SftpLogger: Authentications that can continue: keyboard-interactive,password
10:18:56 [jEdit I/O #2] [message] SftpLogger: Next authentication method: keyboard-interactive
10:18:56 [jEdit I/O #2] [message] SftpLogger: Disconnecting from sles10x64 port 22
10:18:56 [jEdit I/O #2] [error] BrowserIORequest: java.io.IOException: com.jcraft.jsch.JSchException: Auth fail
10:18:56 [jEdit I/O #2] [error] BrowserIORequest: at ftp.SFtpConnection.(SFtpConnection.java:105)
10:18:56 [jEdit I/O #2] [error] BrowserIORequest: at ftp.ConnectionManager.getConnection(ConnectionManager.java:320)
10:18:56 [jEdit I/O #2] [error] BrowserIORequest: at ftp.FtpVFS.getConnection(FtpVFS.java:451)
10:18:56 [jEdit I/O #2] [error] BrowserIORequest: at ftp.FtpVFS._listFiles(FtpVFS.java:218)
10:18:56 [jEdit I/O #2] [error] BrowserIORequest: at org.gjt.sp.jedit.browser.BrowserIORequest.listDirectory(BrowserIORequest.java:159)
10:18:56 [jEdit I/O #2] [error] BrowserIORequest: at org.gjt.sp.jedit.browser.BrowserIORequest.run(BrowserIORequest.java:90)
10:18:56 [jEdit I/O #2] [error] BrowserIORequest: at org.gjt.sp.util.WorkThread.doRequest(WorkThread.java:213)
10:18:56 [jEdit I/O #2] [error] BrowserIORequest: at org.gjt.sp.util.WorkThread.doRequests(WorkThread.java:180)
10:18:56 [jEdit I/O #2] [error] BrowserIORequest: at org.gjt.sp.util.WorkThread.run(WorkThread.java:154)
10:18:56 [jEdit I/O #2] [error] ErrorListDialog$ErrorEntry: sftp://vswfl@sles10x64:22/:
10:18:56 [jEdit I/O #2] [error] ErrorListDialog$ErrorEntry: Cannot list directory: java.io.IOException: com.jcraft.jsch.JSchException: Auth fail

The solution was to modify the sshd_config in /etc/ssh on the sles10 server:
and replaced:

PasswordAuthentication no

with

PasswordAuthentication yes

After this change the SFTP-login and file system browsing in jEdit worked.

Netbeans 7.0 beta is read for download

2010-11-20T10:16:00.000+01:00

Netbeans 7.0beta is ready for download.

I installed on OpenIndiana with jdk 6U22.

The plugin repository contain the python packages.
The installation took very long on my system for the jython core package.

Using lookbehind Regular Expression for searching in jEdit

2010-10-28T17:49:00.002+02:00

Im searching for a string in a logfile, but the string shouldn't be preceded by
two other strings:

($<!(word1|word2)\|)search

The search string is search, which should not preceded
by word1 or word2 followed by a pipe and than the search string.
When there are special characters in the strings, they have to be
escaped.

Installed NetBeans 7.0 Milestone 2 on OpenIndiana Build 147

2010-10-19T09:55:00.000+02:00

I've installed NetBeans 7.0 Milestone 2 on my OpenIndiana Installation Build 147 without problems.

The Python-Module is available from the repository.

PMD, FindBugs, CheckStyle from https://sqe.dev.java.net/ can not be
installed, because it requires maven > 1.0, and 7.0 Milestone 2 installs
maven 2.0, something with the check is wrong.

Sending datablock in Python with > 65536 bytes fails

2010-10-15T13:22:00.000+02:00

In a Simulation Script, I used the ThreadingTCPServer from SocketServer.

The simulation data which should be sent back is larger than 65536 bytes,
but the other side just hang around and doesn't receive everything.

Both parts are executed on the same machine with Solaris 10.

To get all the data over the network, I'm now using a loop sending just
blocks of 10000 bytes and sleeping between with time.sleep( 0.3 ) some time.

I think, the network buffer of the OS isn't large enough to hold the full data
and therefore drops some of it, causing the missing parts on the receiver side.

Resolving merge conflicts in mercurial

2010-10-11T17:13:00.000+02:00

In case of a merge conflict, mercurial gives some hint about it.
You can view the state with the command:

hg resolve -l

If you've resolved a conflict, you have to tell mercurial about this fact:

hg resolve -m file/which/is/resovled

Mercurial has an option to work with multiple patch queues

2010-10-11T17:07:00.000+02:00

Today I encountered the command qqueue in Mercurial.
I haven't found a useful documentation about qqueue except here.

The command allows you to have mutlipel patch queues in a repository,
between which you can switch.

So having a set of patches in one queue and another set of patches in
a different queue.

Soccer Cake

2010-10-02T07:48:00.000+02:00

Debugging memory leaks with DTrace and mdb

2010-09-28T16:11:00.002+02:00

Some time ago I wrote about debugging memory problems with libumem.

Now I've a different approach using dtrace and mdb to find the source code position.

First a let the program execute and finish running under dtrace.
The dtrace-Script is from http://blogs.sun.com/sanjeevb/.

# cat ./memleak.d

#!/usr/sbin/dtrace -s

pid$1:libc.so.1:malloc:entry
{
        self->trace = 1;
        self->size = arg0;
}
pid$1:libc.so.1:malloc:return
/self->trace == 1/
{
        printf("Ptr=0x%p Size=%d", arg1, self->size);
        ustack();
        self->trace = 0;
        self->size = 0;
}

pid$1:libc.so.1:realloc:entry
{
        self->trace = 1;
        self->size = arg1;
        self->oldptr = arg0;
}

pid$1:libc.so.1:realloc:return
/self->trace == 1/
{
        printf("Ptr=0x%p Oldptr=0x%p Size=%d", arg1, self->oldptr, self->size);
        ustack();
        self->trace = 0;
        self->size = 0;
}

pid$1:libc.so.1:calloc:entry
/self->trace == 1/
{
        self->trace = 1;
        self->size = arg1;
}

pid$1:libc.so.1:calloc:return
/self->trace == 1/
{
        printf("Ptr=0x%p Size=%d", arg1, self->size);
        ustack();
        self->trace = 0;
        self->size = 0;
}

pid$1:libc.so.1:free:entry
{
        printf("Ptr=0x%p ", arg0);
}

The result output from the dtrace-Script is postprocessed with a
Perl-Script checking if alloced memory is freed again:

#!/usr/bin/perl

# findleaks.pl

use Data::Dumper;

my %hash = ();

while (<>) {
        if ((/malloc:return Ptr=([^ ]*) Size=(.*)/) || 
            (/calloc:return Ptr=([^ ]*) Size=(.*)/)) {
                $hash{$1} = { size => $2 };
  while (<>) {
   last if /^$/;
   $hash{$1}->{stack} .= $_;
  }
        }
        elsif (/free:entry Ptr=([^ ]*)/) {
                if (exists $hash{$1} and $hash{$1}) {
                        $hash{$1} = '';
                }
        }
 elsif (/realloc:return Ptr=([^ ]*) Oldptr=([^ ]*) Size=(.*)/) {
  if ($1 eq $2) {
   if (exists $hash{$1} and $hash{$1}) {
    $hash{$1} = { size => $3 };
    $hash{$1}->{stack} = '';
    while (<>) {
                          last if /^$/;
                           $hash{$1}->{stack} .= $_;
                  }
   } 
  } else {
    $hash{$1} = '';
    $hash{$2}= { size => $3 };
    $hash{$2}->{stack} = '';
    while (<>) {
                          last if /^$/;
                           $hash{$2}->{stack} .= $_;
                  }
  }
 }

}

foreach my $key (keys %hash) {
 next if not $hash{$key}->{size};
 print "Ptr=$key Size=", $hash{$key}->{size}, "\n";
 print $hash{$key}->{stack}, "\n---------\n";
}

Now we get a list of memory allocations which are not freed after closing
the application.

Such an entry might look like

---------
Ptr=0x81c64f8 Size=20
              libumem.so.1`malloc+0x3e
              libstdc++.so.6`_Znaj+0x1b
              libxerces-c.so.28`_ZN11xercesc_2_818IconvLCPTranscoder9transcodeEPKc+0xc4
              libxerces-c.so.28`_ZN11xercesc_2_89XMLString9transcodeEPKc+0x22
              libvsEAI.so404`readFunction+0x39
              libvsEAI.so404`bcfEAIRepositoryInit+0x5a9
              libvsopCsiBdOOPool.so404`ctrl_call_bcf+0x106
              libvsopCsiBdOOPool.so404`ctrl_call_list+0x1bc
              libvsopCsiBdOOPool.so404`ctrl_gosub_list+0xb2
              libvsopCsiBdOOPool.so404`ctrl_call_bcf+0x7dd
              libvsopCsiBdOOPool.so404`ctrl_call_list+0x1bc
              libvsopCsiBdOOPool.so404`ctrl_gosub_list+0xb2
              libvsopCsiBdOOPool.so404`ctrl_call_bcf+0x7dd
              libvsopCsiBdOOPool.so404`ctrl_call_bcf+0x795
              libvsopCsiBdOOPool.so404`ctrl_call_bcf+0x795
              libvsopCsiBdOOPool.so404`ctrl_call_list+0x1bc
              libvsopCsiBdOOPool.so404`prog_control+0x9a
              progvs_prog`vsop_main+0x5a2
              libvszwflx2Bd.so404`main+0x196
              progvs_prog`_start+0x80

To get the position in the readFunction method, make a core of the application by using

gcore [-F]

and start in mdb with mdb

mdb <core file name >
> libvsEAI.so404`readFunction::dis
libvsEAI.so404`readFunction:    pushl %ebp
libvsEAI.so404`readFunction+1: movl   %esp,%ebp
libvsEAI.so404`readFunction+3: pushl %edi
libvsEAI.so404`readFunction+4: pushl %esi
libvsEAI.so404`readFunction+5: pushl %ebx
libvsEAI.so404`readFunction+6: call   +0x5
libvsEAI.so404`readFunction+0xb:popl   %ebx
libvsEAI.so404`readFunction+0xc:addl   $0x1527d,%ebx
libvsEAI.so404`readFunction+0x12:       subl   $0xd8,%esp
libvsEAI.so404`readFunction+0x18:       leal   0xfffef969(%ebx),%eax
libvsEAI.so404`readFunction+0x1e:       pushl %eax
libvsEAI.so404`readFunction+0x1f:       movl   0x8(%ebp),%edi
libvsEAI.so404`readFunction+0x22:       call   -0x620e
libvsEAI.so404`readFunction+0x27:       movl   %eax,-0x88(%ebp)
libvsEAI.so404`readFunction+0x2d:       leal   0xfffef7a1(%ebx),%eax
libvsEAI.so404`readFunction+0x33:       pushl %eax
libvsEAI.so404`readFunction+0x34:       call   -0x6220
libvsEAI.so404`readFunction+0x39:       movl   %eax,-0x8c(%ebp)
libvsEAI.so404`readFunction+0x3f:       leal   0xfffef972(%ebx),%eax
libvsEAI.so404`readFunction+0x45:       pushl %eax
libvsEAI.so404`readFunction+0x46:       call   -0x6232
libvsEAI.so404`readFunction+0x4b:       movl   %eax,-0x90(%ebp)
libvsEAI.so404`readFunction+0x51:       leal   0xfffef978(%ebx),%eax
libvsEAI.so404`readFunction+0x57:       pushl %eax
libvsEAI.so404`readFunction+0x58:       movl   %eax,-0xb0(%ebp)
libvsEAI.so404`readFunction+0x5e:       call   -0x624a
libvsEAI.so404`readFunction+0x63:       leal   0xfffef97b(%ebx),%edx

Because of the offset from the stack trace libvsEAI.so404`readFunction+0x39 we see in the disassembly, that the second call to libxerces-c.so.28`_ZN11xercesc_2_89XMLString9transcodeEPKc+0x22 is the problem.

The method called is

>echo "_ZN11xercesc_2_89XMLString9transcodeEPKc" | c++filt
xercesc_2_8::XMLString::transcode(char const*)

SFTP withing jedit doesn't work in OpenIndiana

2010-09-21T14:38:00.000+02:00

I use a lot the FTP plugin from jEdit.
This plugin allows you to open a SFTP connection to a remote host via ssh and
work with files from the remote host, as if they are local. The plugin is
available via the file browser.

You can load and save the files.

Within OpenIndiana this just doesn't work anymore.
The activity.log shows:

8:39:58 AM [jEdit I/O #3] [message] SftpLogger: Remote version string: SSH-2.0-Sun_SSH_1.0.1
8:39:58 AM [jEdit I/O #3] [message] SftpLogger: Local version string: SSH-2.0-JSCH-0.1.43
8:39:58 AM [jEdit I/O #3] [message] SftpLogger: CheckCiphers: aes256-ctr,aes192-ctr,aes128-ctr,aes256-cbc,aes192-cbc,aes128-cbc,3des-ctr,arcfour,arcfour128,arcfour256
8:39:58 AM [jEdit I/O #3] [message] SftpLogger: aes256-ctr is not available.
8:39:58 AM [jEdit I/O #3] [message] SftpLogger: aes192-ctr is not available.
8:39:58 AM [jEdit I/O #3] [message] SftpLogger: aes256-cbc is not available.
8:39:58 AM [jEdit I/O #3] [message] SftpLogger: aes192-cbc is not available.
8:39:58 AM [jEdit I/O #3] [message] SftpLogger: arcfour256 is not available.
8:39:58 AM [jEdit I/O #3] [message] SftpLogger: SSH_MSG_KEXINIT sent
8:39:58 AM [jEdit I/O #3] [message] SftpLogger: SSH_MSG_KEXINIT received
8:39:58 AM [jEdit I/O #3] [message] SftpLogger: kex: server->client aes128-cbc hmac-md5 none
8:39:58 AM [jEdit I/O #3] [message] SftpLogger: kex: client->server aes128-cbc hmac-md5 none
8:39:58 AM [jEdit I/O #3] [message] SftpLogger: SSH_MSG_KEXDH_INIT sent
8:39:58 AM [jEdit I/O #3] [message] SftpLogger: expecting SSH_MSG_KEXDH_REPLY
8:39:59 AM [jEdit I/O #3] [message] SftpLogger: ssh_rsa_verify: signature false
8:39:59 AM [jEdit I/O #3] [message] SftpLogger: Disconnecting from onbahaspa port 22
8:39:59 AM [jEdit I/O #3] [error] BrowserIORequest: java.io.IOException: com.jcraft.jsch.JSchException: verify: false
8:39:59 AM [jEdit I/O #3] [error] BrowserIORequest: at ftp.SFtpConnection.(SFtpConnection.java:105)
8:39:59 AM [jEdit I/O #3] [error] BrowserIORequest: at ftp.ConnectionManager.getConnection(ConnectionManager.java:320)
8:39:59 AM [jEdit I/O #3] [error] BrowserIORequest: at ftp.FtpVFS.getConnection(FtpVFS.java:451)
8:39:59 AM [jEdit I/O #3] [error] BrowserIORequest: at ftp.FtpVFS._canonPath(FtpVFS.java:194)
8:39:59 AM [jEdit I/O #3] [error] BrowserIORequest: at org.gjt.sp.jedit.browser.BrowserIORequest.listDirectory(BrowserIORequest.java:158)
8:39:59 AM [jEdit I/O #3] [error] BrowserIORequest: at org.gjt.sp.jedit.browser.BrowserIORequest.run(BrowserIORequest.java:90)
8:39:59 AM [jEdit I/O #3] [error] BrowserIORequest: at org.gjt.sp.util.WorkThread.doRequest(WorkThread.java:213)
8:39:59 AM [jEdit I/O #3] [error] BrowserIORequest: at org.gjt.sp.util.WorkThread.doRequests(WorkThread.java:180)
8:39:59 AM [jEdit I/O #3] [error] BrowserIORequest: at org.gjt.sp.util.WorkThread.run(WorkThread.java:154)
8:39:59 AM [jEdit I/O #3] [error] ErrorListDialog$ErrorEntry: sftp://hartter@onbahaspa:22/~/:
8:39:59 AM [jEdit I/O #3] [error] ErrorListDialog$ErrorEntry: Cannot list directory: java.io.IOException: com.jcraft.jsch.JSchException: verify: false

The verification fails. I had written to the OpenIndiana Mailinglist and

the successful suggestion from Onno Molenkamp was, to remove the

sun.security.pkcs11.SunPKCS11 provider from /usr/jdk/latest/jre/lib/security/java.security.

After removing and renumbering the providers there, the SFTP Plugin works again in OpenIndiana.

Upgraded to OpenIndiana 147 from OpenSolaris Build 134

2010-09-16T09:33:00.001+02:00

Now I upgraded from OpenSolaris Build 134 to OpenIndiana Build 147.

Because just testing in VirtualBox fails. The ISO boots and I can install the
system, but during the startup from disk, the kernel likes to update the
microcode and fails.

The upgrade works as usual see the Wiki for more information.

I had the same problem as describe in the wiki with the download of
thunderbird.

Just uninstalled thunderbird and added the strange hostname
~0~pkg-origin.openindiana.org with an IP from nslookup to my local hosts files.

After this the upgrade worked with some repetitions because of Error 503 Service unavailble.

After restarting, I didn't get my german keyboard layout.

After searching, I rememberd the fact, that Build 145 changed the keyboard
to be a SMF service.
See the BugIusse @ Illumos

# svccfg
svc:> select keymap:default
svc:/system/keymap:default> setprop keymap/layout = German

svc:/system/keymap:default> refresh

svc:/system/keymap:default> exit

Changing the message of a mercurial Patch before commiting into the repository

2010-09-10T07:59:00.000+02:00

When using Mercurial Queues to work with patches, you can set a message,
when starting a patch.

When you like to change the commit message, when you qfinish the patch,
you have no chance currently.

The only way is to just push the patch you like.
Prepare a file or add the text directly to the qrefresh command.

hg qrefresh -l /path/to/msgfile

hg qrefresh -e "my commit text"

Python pygtk based MergeTool for Mercurial

2010-09-09T09:41:00.000+02:00

I've found a merge tool named gpyfm, which is a rewrite of the old TeamWare file merge tool used by Sun Engineers.

Excerpt from the Homepage:

gpyfm was originally written by David Marker (assitant ON Nevada gatekeeper). Richard Lowe, one of the project leads for SCM Migration.

Here is a picture from the tutorial of the merge tool:

Compiling Qt 4.5.3 on OpenSolaris

2010-09-08T16:19:00.000+02:00

Because the tortoisehg is changing to PyQt instead of PyGtk, I attempted to
compile the Qt Package from Nokia ( former Trolltech ).

I used the Sun Studio Compilers, but ran into a compile problem, where I had
to set an additional compile/link flag see here for the message.

The configure option I used was:

CC=/opt/sunstudio12.1/bin/cc CXX=/opt/sunstudio12.1/bin/CC CXXFLAGS="-library=stlport4" LDFLAGS="-library=stlport4" ./configure --prefix=/tools/qt/4.5.3 -opensource

After this, just a gmake and gmake install and everything was fine.

Moving source changes to a previous patch in Mercurial with Mercurial Queues ( MQ )

2010-09-03T14:52:00.000+02:00

I started to use Mercurial Queues to collect my changes in a patch.
Now I have more patches applied in my local repository.

When detecting an error, I changed the error in the code, but it belongs
to an earlier patch than the topmost which is now active.

To get the changes there, I do now

hg diff > /tmp/patch
hg qgoto targetpatch
hg import --no-commit /tmp/patch
hg qrefresh

This allows me to move source changes to a different patch .

Getting information in screen when something happens on an other shell

2010-09-02T10:52:00.000+02:00

Some time ago, I started to use screen in my terminal instead of opening
multiple registers or gnome-terminals. See my starting entry here.

Now, I've running some simulation scripts which receive messages. When you
are on a different shell in screen and something happens on an other screen-shell, you can get notified by a special key sequence.

Go the the shell in screen, where you like to receive notifications and press

-

After this you get a hint, when something changes in the output.

John Deere Cake

2010-08-29T16:41:00.000+02:00

Peach Cake

2010-08-29T10:08:00.000+02:00

Bottle Cake

2010-08-29T08:53:00.000+02:00